Skip to content

[Bug]: Dataflow keeps retrying to obtain an already deleted BigQuery dataset #28438

Description

@fozzie15

What happened?

max retry is not applying to getDatasetLocation() and attempts to retry indefinitely.

Max retries is seen in the stack trace here at 9, but beam does not let go after the 9 retries, and the logs are filled with continuous attempts to getDatasetLocation (see below.) This is a singular instance of the stack trace that repeats indefinitely, hanging the job:

Error message from worker: java.lang.RuntimeException: unable to obtain dataset for dataset nonexistent_dataset in project <project_id>
org.apache.beam.sdk.io.gcp.bigquery.BigQueryHelpers.getDatasetLocation(BigQueryHelpers.java:598)
org.apache.beam.sdk.io.gcp.bigquery.WriteTables.startLoad(WriteTables.java:506)
org.apache.beam.sdk.io.gcp.bigquery.WriteTables.access$700(WriteTables.java:97)
org.apache.beam.sdk.io.gcp.bigquery.WriteTables$WriteTablesDoFn.processElement(WriteTables.java:277)
Caused by: java.io.IOException: Unable to get dataset: nonexistent_dataset, aborting after 9 retries.
org.apache.beam.sdk.io.gcp.bigquery.BigQueryServicesImpl.executeWithRetries(BigQueryServicesImpl.java:1476)

Issue Priority

Priority: 2 (default / most bugs should be filed as P2)

Issue Components

  • Component: Python SDK
  • Component: Java SDK
  • Component: Go SDK
  • Component: Typescript SDK
  • Component: IO connector
  • Component: Beam examples
  • Component: Beam playground
  • Component: Beam katas
  • Component: Website
  • Component: Spark Runner
  • Component: Flink Runner
  • Component: Samza Runner
  • Component: Twister2 Runner
  • Component: Hazelcast Jet Runner
  • Component: Google Cloud Dataflow Runner

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions